Jh 3117241732

7/29/2019 Jh 3117241732

http://slidepdf.com/reader/full/jh-3117241732 1/9

Mr. Siddique Ahmed, Mr. Gajendra Patil / International Journal of Engineering Research and

Applications (IJERA) ISSN: 2248-9622 www.ijera.com

Vol. 3, Issue 1, January -February 2013, pp.1724-1732

1724 | P a g e

Bumper Shape Optimization for Pedestrian Safety

Mr. Siddique Ahmed*, Mr. Gajendra Patil**

*( Asst. Professor in Mechanical Department, SIES Graduate School Of Technology, Navi Mumbai,India,)

** (Asst. Professor in Mechanical Department,, Pillai's Institute of Information Technology, Navi MumbaiIndia,)

Abstract:Almost 28% of the total accidents are

encountered annually in road traffic crashes

worldwide are pedestrians. Pedestrian means a

person traveling on foot,

whether walking or running. The objective of

project is to Optimize Bumper shape to increase

pedestrian safety. The main focus will be to

design a bumper shape to reduce lower andupper leg injuries. The system will be analyzed

using computational codes like LS Dyna andOptimization tools like HyperStudy. This study

indicates that the increase in the bumper area to

optimize its shape not always a solution. If we just

keep on increasing the bumper size then the

bumper becomes very weak and fails during the

pedestrian collision.The bumper shape

optimization is done by a new method called

HPERMORPH technique.

Keywords: LSDYNA, HYPERMORPH,HYPERMESH

1. INTRODUCTION:Almost 28% of the total accidents are

encountered annually in road traffic crashesworldwide are pedestrians. Pedestrian means a person traveling on foot,

whether walking or running. In some communities,those traveling using roller skates or skateboards arealso considered to be pedestrians. In modern times,the term mostly refers to someone walking ona road or footpath.

Figure 1: Contribution of pedestrian in totalaccident [1]

Despite the magnitude of the problem, most attempts

at reducing pedestrian deaths have focused solely oneducation and traffic regulation. However, in recentyears crash engineers have begun to use design

principles that have proved successful in protectingcar occupants to develop vehicle design conceptsthat reduce the likelihood of injuries to pedestriansin the event of a car-pedestrian crash. These involve

redesigning the bumper, hood (bonnet), and the

windshield and pillar to be energy absorbing (softer)without compromising the structural integrity of the

car. Most pedestrian deaths occur due to thetraumatic brain injury resulting from the hard impactof the head against the stiff hood or windshield. Inaddition, although usually non-fatal, injuries to the

lower limb (usually to the knee joint and long bones)are the most common cause of disability due to pedestrian crashes.

2. KINEMATICS OF THE

PEDESTRIAN ACCIDENTIn order to improve the crash performance

of a car’s front end concerning pedestrian protection,the first important step is to analyze the kinematicsthe human body experience in a car impact. Thus the

parameter that influence on one hand the pedestrian’s kinematics and on the other hand thecar’s crash behaviour can be investigated. Thisexamination allows localizing critical impact zonesthat need consideration regarding pedestrian safety.Information about the kinematics of the human body

in case of an impact on the car’s front end can beobtained by evaluation of accident data, byreconstruction of accidents, furthermore by cadaver and dummy tests and in additional by simulation

using a full dummy model. To perform simulation of full dummy impact a rigid body dummy is integratedinto the finite element model of the vehicle. [7]The examples show the impact of a 6 years old child

and adult male on a sport car and on a van. Theanimation illustrates that the kinematics of the pedestrian are depending on one hand of the

pedestrian’s size and weight and on the other handon the car’s front structure. In the first contact of the6 years old child to the car, many body parts are

involved: such as upper leg, pelvis and torso areimpacted by the bumper area. In the case of van, an

even larger area is impacted due to the bigger bumper system of this vehicle. In the next step, the

child’s head hits the forward section of the bonnet

http://www.ijera.com/



7/29/2019 Jh 3117241732





1725 | P a g e

top. In the case of the adult male, the first contact iswith the leg by the vehicle’s bumper system,initiating a rotation of the whole body. Depending

on the car’s front structure, the pelvis hits the bonnetleading edge; in this case the relatively flat sport car,

pelvis and car do not get into contact whereas in caseof the van, the upper leg and pelvis hit the bonnettop (as in the sports car) or the windscreen area (asin the van).

Figure 2: A 6 years old child and an adult male

hit by a sport car [7]

Figure 3: A 6 years old child and an adult male

hit by a van [7]

Thus concerning adults, the collision of the pedestrian with the vehicle can be divided into three

impact phases:

Bumper hits the leg; rotation of the body isinitiated.

Pelvis hits the bonnet leading edge(depending on the vehicle)

Head hits the bonnet top or the

windscreenReduced Car Model of ToyotaYaris and Analysis of the Original model

3. REDUCED CAR MODEL OF TOYOTA

YARIS AND ANALYSIS OF

ORIGINAL MODEL

Figure 4: reduced Toyota Yaris

3.1 Reduced FE Model Preparations

FE Model Quality:The FE model quality as measured in

Hypermesh is given in table below.

Sr. No.

ElementProperty

PermissibleValue

ActualValue

1MinimumElement Size

5 5.7

2 Warpage < 15 15

3Aspect Ratio

<5 2.13

4 Skew Angle < 60 44.09

5JacobeanRatio >

0.6 0.62

6Min Angle(Quad) >

35 36.74

7Max Angle

(Quad) <140 138.07

8Min Angle

(Tria) >20 34.32

9Max Angle(Tria) <

120 99.4

10 % of Trias /Pentas < 10% 7%

Table 1: FE Model Quality CriteriaThe full co related model of Toyota Yaris is

reduced up to the firewall section. The engineassembly, suspension and the wheels are excludedfrom model as they do not contribute to the bumper stiffness. The reduced model helps in fast processing

of the LS Dyna runs due to reduced number of elements. The reduced model is validated for stiffness and connection by performing a modal

analysis. The FE model considered is as shown

below.




7/29/2019 Jh 3117241732





1726 | P a g e

Figure 5: FE model of Toyota Yaris

3.2 FE Model SetupThe analysis setup consists of reduced car

model and lower legform. The vehicle is placed atground level and legform impactor is placed at 25mm above ground level.

Figure 6: Impact test setup

3.3 Regulatory Requirements according to

Automotive Industry Standard 100

The maximum dynamic knee bending angleshall not exceed 19°.

The maximum dynamic knee shearing

displacement shall not exceed 6.0 mm. The acceleration measured at the upper end of

the tibia shall not exceed 170 g.

Figure 7: Pedestrian leg-form criterion[8]

Figure 8: Lower part of human body [9]

FEA TOOLSThe following software’s are utilized for

Pre Processing (FEA model discritization), solving

(FEA computation code) and Post Processing (FEAresult visulization)

Pre Processor - Hypermesh 9.0Solver -LS DynaPost Processor -Hyperworks, LS Prepost

RESULTS FOR (ORIGINAL MODEL)

BASELINE ITERATION: The Maximum Tibia acceleration is 217.9 g

which is higher than the allowable limit of

170g. Hence, baseline design does not meetacceleration requirement.




7/29/2019 Jh 3117241732





1727 | P a g e

The Maximum Knee Shearing displacementis 2.87mm which is within allowable limit

of 6 mm. Hence meets the requirement.

The Maximum Knee Bending angleobserved is 20.84 deg. which is higher than

the allowable limit of 19 deg. Hence, baseline design does not meet accelerationrequirement.

MODIFICATION OF BUMPER SHAPE

FOR OPTIMIZED RESULTS:The different shapes of bumper are created

by using the HyperMorph module in HyperMesh.The HyperMorph module allows you to alter modelsin useful, logical, and intuitive ways while keepingmesh distortion to a minimum. Morph volume

technique is used to modify bumper shapes. Thismethod is ideal for making simple changes to

complex models. This method encloses the mesh inone or more deformable 3-D blocks. Each block is a"morph volume" (often referred to as "mvol" for brevity) and governs the movement of the meshwithin its boundaries. Thus, by changing the shapesof the blocks, you can change the shape of the mesh.The process of morphing consists of entering either

the domains panel to create global or local domainsand handles, the morph volumes panel to createmorph volumes, or the freehand or map to geom panels to morph the mesh directly. Morphing using

domains, handles, and morph volumes is done in themorph panel by moving the handles and in the mapto geom panel by mapping domains, nodes,elements, or morph volume edges to geometric

entities. The freehand panel can be used with modelsthat have (or lack) domains, handles, and morphvolumes.

The red dots indicate the morph handles. Thehandles can be translated to get desired change inshape. We have altered the handles in specific

manner to modify the bumper shape incrementally.In all six combinations of shape were made byaltering the dimensions of the handles.

Figure 9: Morphing handle notation

Table 2: Showing Details for Handle Movement

for Shape GenerationThe handles were moved in X direction as shown in

figure to get six different shapes of bumper.The six shapes obtained are as given below.

Figure 10: Six different shapes

RESULTS FOR SHAPE OPTIMIZATION:

1.1. Iteration for Shape 3

1.1.1. Animation Instances




7/29/2019 Jh 3117241732





1728 | P a g e

1.1.2. Stress and Strain Plots

Figure 11: Von mises and plastic strain plots

1.1.3. Legform Outputs

Figure 12: Knee Bending Angle

Figure 13: Upper Tibia Acceleration and Knee

Shearing Displacement Plots






7/29/2019 Jh 3117241732





1729 | P a g e

1.2.2. Stress and Strain Plot


1.2.3. Legform Outputs





7/29/2019 Jh 3117241732





1730 | P a g e





1.3.2. Stress and Strain Plots


1.3.3.

Legform Outputs




7/29/2019 Jh 3117241732





1731 | P a g e




RESULT SUMMARY & OBSERVATIONS

Table 3: Result Summary

From the above study following observations arededuced

Baseline Design: Only meet Knee shearingdisplacement and shows maximum upper tibiaacceleration among all shape changes i.e.

217g which is beyond acceptable limit of 170g.

Shape 01: Marginal improvement over the baseline design but shows same failure trend

compared to Baseline Design.

Shape 02: Improved over earlier designshapes, only does not meet Upper Tibia

Acceleration. Shape 03: Bumper behavior is worsen

compared to earlier shape. Also shows PlasticStrain beyond allowable limits.

Shape 04: All the requirements are satisfied

and are well bellow the acceptable limits withensuring the component integrity.

Shape 05: Component meet the regulatoryrequirement but fails in component integrityas plastic strain shows values beyondallowable limit f 22%.

Shape 06: Bumper behavior has worsened ascompared to earlier shape. Also shows PlasticStrain beyond allowable limits.

CONCLUSION

Shape 04 is the best possible shape among alldesign as all the requirements are satisfiedand are well below the acceptable limits with

ensuring the component integrity.

The Max. Tibia acceleration is 126.69 g

which is lower than the allowable limit of 170g. Hence, meet acceleration requirement.

The Max. Knee Shearing displacement is 2.68

mm which is within allowable limit of 6 mm.Hence meets the requirement.

The Max. Knee Bending angle observed is11.51 deg. which is higher than the allowablelimit of 19 deg. Hence, baseline design does

not meet acceleration requirement.

Energy transfer has a vital role in shapeoptimization.

To reduce upper tibia acceleration designer need to avoid Sharpe changes in the Bumper at first hitting area, which is well evident from

first three shape changes.

Increase in the bumper area to optimize its

shape not always a solution which is wellevident from shape 05 & 06.

4. RECOMMENDATIONS The Bumper optimization is studied on

reduced model due to computationallimitation of available software.

The material data used is just three pointcurve data in nonlinear zone. Detailedmaterial curve will give the good results.

Bumper is modeled with 5mm minimum

element size. Fine model will help us to getthe more accurate prediction of knee angle

and acceleration.




7/29/2019 Jh 3117241732





1732 | P a g e

Material change is not considered in the bumper designing.

Energy absorptions techniques need to be

also considered while designing bumper shape such as use of soft and hard foam.

REFERENCE:1. Andy Bailey, Koki Ikeda, Hideki Ishitobi,

“Development Of Aluminium HoodStructure For Pedestrian Protection”,

Toyota Motor Corporation,2. D.K. Park, C.D.Jang, S.B. lee, S.J. Heo,

H.J. Yim, M.S. Kim. (2010), “Optimizing

The Shape Of A Bumper Beam SectionConsidering Pedestrian Protection”,International Journal of Automotive

Technology, Vol. 4, pp.489-4943. T.L. teng, V.L NGO (2011), “analyzing

Pedestrian Head Injury To DesignPedestrian – Friendly Hood”, InternationalJournal of Automotive Technology, Vol.12, pp.213-224

4. A.Abvabi, A.Nasr, A.Noorpoor, M.S.

Kiasat (2009), “Lower Extremity Injuries Invehicle – pedes trian Collisions Using Alegform Impactor Model”, Journal of

Zhejiang University, Vol. 2, pp.97-1055. Tso-Liang Teng , Trung – Kien Le (2009),

“Development and validation Of aPedestrian Deformable Finite Element

Model”, Journal of Mechanical Science

And Technology, Vol.23 . pp.2268-22766. Giovanni Belingardi, Ermias Gebrekidan

Koricho, Brunetto Martorana.(2011),“Design Optimization And Implementationof Composite And RecyclableThermoplastic Materials For AutomotiveBumper”, Fifth International ConferenceOn Advanced computational Methods in

Engineering7. Susanne Dorr, Hartmut Chladek, Armin

Hub, “Crush Simulation in PedestrianProtection”, 4th European LS-DYNA Users

Conference.8. Jae Wan Lee, Youn Soo Kang, Gyung-Jin

Park., “Vehicle Hood and Bumper Structure Design To Mitigate casualties Of

Pedestrian accidents”9. Wikipedia, the free encyclopedia (last

visited on 1st march 2012)

10. Peter J. Schuster. (2006), “Current Trendsin Bumper Design For Pedestrian Impact”,California Polytechnic State University,U.S

11. Svoboda Jiri, Kuklik Martin (2006),“Influence of Bumper design to Lower Legimpact Response” , Czech Technical

University in Prague, Czech Republic




Documents

Jh 3117241732